To analyze the risk factors of retinopathy of prematurity (ROP) and provide evidence for the rational establishment of screening standard. The clinical data of 1675 preterm infants at gestational age < or = 36 weeks or birth weight < or = 2500 g who were admitted to the neonatal intensive care unit and had been screened in our hospital from July 2006 to May 2008 were analyzed retrospectively by univariate analysis and Logistic regression analysis. Gender, birth count, gestational age, birth weight, oxygen therapy, and mother's conditions were recorded. ROP was detected in 195 (11.6%) of 1675 infants, of whom 35 infants (2.1%) had type 1 or threshold ROP. The lower the birth weight, the smaller the gestational age and the longer the time of oxygen therapy were, the higher the incidence of ROP was. For the infants whose birth weight was < or = 1200 g, 1201 - 1500 g, 1501 - 2000 g, 2001 - 2500 g, the incidence of ROP was 73.2%, 30.4%, 8.0%, and 1.1%; for those at gestational age < or = 30 weeks, 30(+1)-32 weeks, 32(+1)-34 weeks, 34(+1)-36 weeks, the incidence of ROP was 67.6%, 16.9%, 3.9%, and 1.0%; for the infants underwent oxygen therapy for 0 d, -3 d, -5 d, -8 d, > 8 d, the incidence of ROP was 1.5%, 3.3%, 9.6%, 23.2% and 38.8%;in the infants who inhaled oxygen at concentrations of 0.40, -0.60, -0.80 and > 0.80, the incidence of ROP was 11.8%, 18.1%, 26.8%, and 52.6%, respectively. Logistic regression analysis indicated that low birth weight, small gestational age, asphyxia, apnea, oxygen therapy were the high risk factors of ROP (the odds ratio was 0.957, 1.052, 1.186, 5.314, and 1.881). Low birth weight, small gestational age, asphyxia, apnea, and oxygen therapy were the high risk factors of ROP. It is recommended that all preterm infants with high risk factors should be screened.